ABSTRACT
Computational modeling of skin color and/or skin reflectance spectra opens up new ways to investigate functional properties of human skin. Modeling of skin color and its variations associated with the physiological changes in human skin, such as blood oxy-and deoxygenation, melanin content, etc., is frequently required in various medical and biomedical applications. We present an open-access computational tool for online simulation of skin color and/ or skin reflectance and transmittance spectra in real time. Human
skin is presented as multi-layered medium. The variations in spatial distribution of blood, pheomelanin, eumelanin, index of blood oxygen saturation, hematocrit, and volume fraction of water are taken into account. The developed Monte Carlo (MC)-based calculator of spectra and color of human skin is supported by Compute Unified Device Architecture (CUDA), introduced by NVIDIA Corporation, that provides acceleration of modeling up to 103 times, allowing produce the results of simulation within seconds. The calculator is based on the object-oriented programming (OOP) paradigm and available online at www.biophotonics.ac.nz.Examples of MC modeling of skin optical properties optimal for removal the tattoo or any other localized absorbing abnormality by laser thermolysis are also presented. This optimization is based on the laser wavelength selection and application of immersion optical clearing for enhancement of laser light selective absorption. 2.1 Introduction
In vivo measurements of human skin spectra serve as an important supplement to standard non-invasive optical techniques for diagnosing various skin diseases [1], such as venous ulcers, skin necrosis, and interstitial edema. However, the quantified analysis of the reflectance spectra is complicated by the fact that skin has a complex multilayered non-homogeneous structure with a spatially varying absorption coefficient, mainly determined by melanin pigmentation, oxygen saturation of cutaneous blood, index of erythema, contents of bilirubin, β-carotene, and other chromophores. Various approaches targeting the modeling of human skin reflectance spectrum and associated colors exist, but in our current work we apply the recently developed multipurpose graphics-processing unit (GPU)-accelerated MC tool for the needs of biophotonics and biomedical optics [2-4].The description of optical radiation propagation within random media is based on the radiative transfer theory [5] that forms a basis of MC modeling of photons migration in biological tissues [6]. Originally introduced in biomedical optics for the counting of fluence rate distribution in biological tissues for the purpose of estimation laser radiation dose [7], in the last decades the MC approach has become a primary tool for a number of needs in biomedical optics. Incorporated with the computational model of
human skin [8] MC technique has been used for simulation of skin visual and near-infrared reflectance spectra [9,10], analysis of skin fluorescence excitation [11-13], simulation of optical coherence tomography (OCT) images of human skin [14,15], analysis of scattering orders, and OCT image formation [16-18]. The MC approach has been generalized for simulation of coherent effects of multiple scattering, such as enhancement of coherent back-scattering (CBS) and changes of temporal intensity correlation function depending on the dynamics of scattering particles [19,20]. Based on these developments a new approach of handling polarization has been introduced and some effects such as a helicity flip of circular polarization has been observed [21,22]. The obtained modeling results have been comprehensively validated by comparison with the known exact solution by Milne [23,24] and with the results of experimental studies of image transfer through the water solution of spherical microparticles of known size and density [25,26]. Meanwhile, a number of other MC algorithms has been developed in the past, see for example [27-30].
